Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <3dac0be8-6a49-43db-da65-e99d2e9719e6@collabora.com>
Date:   Wed, 19 Apr 2023 10:10:58 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.9.1
Subject: Re: [PATCH linux-next v3 4/4] clocksource/drivers/timer-mediatek:
 Make timer-mediatek become loadable module
To:     walter.chang@mediatek.com,
        Daniel Lezcano <daniel.lezcano@linaro.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Matthias Brugger <matthias.bgg@gmail.com>,
        "Maciej W . Rozycki" <macro@orcam.me.uk>,
        John Stultz <jstultz@google.com>
Cc:     wsd_upstream@mediatek.com, stanley.chu@mediatek.com,
        Chun-hung.Wu@mediatek.com, Freddy.Hsin@mediatek.com,
        linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
        linux-mediatek@lists.infradead.org
References: <20230419074910.10809-1-walter.chang@mediatek.com>
 <20230419074910.10809-5-walter.chang@mediatek.com>
Content-Language: en-US
From:   AngeloGioacchino Del Regno 
        <angelogioacchino.delregno@collabora.com>
In-Reply-To: <20230419074910.10809-5-walter.chang@mediatek.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

Il 19/04/23 09:49, walter.chang@mediatek.com ha scritto:
> From: Chun-Hung Wu <chun-hung.wu@mediatek.com>
> 
> Make the timer-mediatek driver which can register
> an always-on timer as tick_broadcast_device on
> MediaTek SoCs become loadable module in GKI.
> 
> Tested-by: Walter Chang <walter.chang@mediatek.com>
> Signed-off-by: Chun-Hung Wu <chun-hung.wu@mediatek.com>

I think I typoed your email when sending the example patch for the
conversion to platform_device. Check [1], it may be better to just
iterate through that? (please ignore the pure_initcall() part, that's
a mistake, it's never gonna happen as it automatically becomes a
module_init() call).

It depends on what maintainers think about that clocksource.h addition,
the patch got zero comments, so if you're interested in that perhaps we
can explicitly ask what would be the best option between yours and mine;
that addition is done only to avoid the big ifdef party that this patch
proposes and makes things a bit shorter if this timer modularization
goes on with more drivers, but I don't have strong opinions anyway.

In the meanwhile, just to eventually speed up integrating this, or the
other patch - I'll still give you a review of this one.

[1]: 
https://patchwork.kernel.org/project/linux-mediatek/patch/20230309132119.175650-1-angelogioacchino.delregno@collabora.com/

> ---
>   drivers/clocksource/Kconfig          |  2 +-
>   drivers/clocksource/timer-mediatek.c | 39 ++++++++++++++++++++++++++++
>   2 files changed, 40 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig
> index 526382dc7482..a7413ad7b6ad 100644
> --- a/drivers/clocksource/Kconfig
> +++ b/drivers/clocksource/Kconfig
> @@ -472,7 +472,7 @@ config SYS_SUPPORTS_SH_CMT
>   	bool
>   
>   config MTK_TIMER
> -	bool "Mediatek timer driver" if COMPILE_TEST
> +	tristate "Mediatek timer driver"

While at it, you could also fix the text, Mediatek -> MediaTek

>   	depends on HAS_IOMEM
>   	select TIMER_OF
>   	select CLKSRC_MMIO
> diff --git a/drivers/clocksource/timer-mediatek.c b/drivers/clocksource/timer-mediatek.c
> index 7bcb4a3f26fb..3448848682c0 100644
> --- a/drivers/clocksource/timer-mediatek.c
> +++ b/drivers/clocksource/timer-mediatek.c
> @@ -13,6 +13,9 @@
>   #include <linux/clocksource.h>
>   #include <linux/interrupt.h>
>   #include <linux/irqreturn.h>
> +#include <linux/module.h>
> +#include <linux/of_device.h>
> +#include <linux/platform_device.h>
>   #include <linux/sched_clock.h>
>   #include <linux/slab.h>
>   #include "timer-of.h"
> @@ -337,5 +340,41 @@ static int __init mtk_gpt_init(struct device_node *node)
>   
>   	return 0;
>   }
> +
> +#ifdef MODULE

#ifndef MODULE
... two lines...
#else
... a bunch of lines ...
#endif

looks more readable. I'd go with that.

> +static int mtk_timer_probe(struct platform_device *pdev)
> +{
> +	int (*timer_init)(struct device_node *node);
> +	struct device_node *np = pdev->dev.of_node;
> +
> +	timer_init = of_device_get_match_data(&pdev->dev);
> +	return timer_init(np);
> +}
> +
> +static const struct of_device_id mtk_timer_match_table[] = {
> +	{
> +		.compatible = "mediatek,mt6577-timer",
> +		.data = mtk_gpt_init,

Fits in one line!

> +	},
> +	{
> +		.compatible = "mediatek,mt6765-timer",
> +		.data = mtk_syst_init,

ditto.

> +	},
> +	{}

Always end with { /* sentinel */ }

> +};
> +
> +static struct platform_driver mtk_timer_driver = {
> +	.probe = mtk_timer_probe,
> +	.driver = {
> +		.name = "mtk-timer",

"mediatek-timer" looks nicer :-)

> +		.of_match_table = mtk_timer_match_table,
> +	},
> +};
> +module_platform_driver(mtk_timer_driver);
> +
> +MODULE_DESCRIPTION("MediaTek Module Timer driver");

"MediaTek Timer driver" is enough, "Module" gets misleading if this gets compiled
as built in platform driver (instead of built in timer_of).

> +MODULE_LICENSE("GPL v2");
> +#else
>   TIMER_OF_DECLARE(mtk_mt6577, "mediatek,mt6577-timer", mtk_gpt_init);
>   TIMER_OF_DECLARE(mtk_mt6765, "mediatek,mt6765-timer", mtk_syst_init);
> +#endif